This paper presents a model predictive control (MPC) approach based on the mixed integer linear programming (MILP) to develop an optimal power management strategy (PMS) for minimizing the electricity bill of commercial buildings in a domestic on-grid system. The optimal PMS is first formulated as a MILP-MPC with time-varying constraints. The constraints are then linearized at each sampling time so that a receding horizon principle can be used to determine the control input applied to the plant

more »

... nd update the model. The time-varying efficiency of power electronic converters is evaluated for each time interval and assumed to be persistent for the prediction time horizon. The numerical results show that the proposed MILP-MPC strategy with variable efficiency is effective in utilizing photovoltaic power generation to save the cost on electricity for buildings.